You can get a decent, 3 bed, 2 bath house for 100k. Just move to some place like Tucumcari, NM. Why not? Oh...right...the same reasons no one else moves to places like that...
Preferences don’t explain why we aren’t building housing where people want to live. Mid rise buildings don’t need to be particularly expensive per square foot. ~11 million for a 50 unit building is 220k / apartment not 100k cheap but way better than what you see near most cites people want to live in. 2 to 3x housing density requires extra transportation infrastructure but it also means being able to support such infrastructure.
Instead walk around most expensive city’s and you see single family dwellings /row houses in sight of high rises / skyscrapers. That’s not economic efficiency that’s people who can afford high housing prices likening the system the way it is.
You need to find a way to convince the owners of that inefficient housing to sell it to a developer so that they can demolish it to build the efficient housing. That will add significantly to the unit economics of the efficient housing.
At the start you might be adding a few million in land costs and building taller, but that quickly deflates the housing market. Pushing people to sell before their homes become ever less valuable. Further cities outlive people, reluctant homeowners eventually die.
City infrastructure similarly has real costs, but an infrastructure tax on every net new unit isn’t going to see anywhere close to current prices.
Sure, you could argue that some people prefer to not live a destitute life, and that influences the high price of housing. But that is reductive, ignoring a host of other factors, which again, you might be able to boil down to preference (wealthy capitalists prefer to make more money) but again, it's reductive and offers a somewhat shallow perspective and not much to act on.
You would at least be able to act on the true cause rather than chase short term changes that may not even work or won't scale. If it's indicative of a distribution problem, then we should be investigating distribution solutions. If you can't see this connection, then I posit that you might have the shallow perspective.
It seems like job location, compensation, average cost of living, and commute would play a fairly large role.